IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
Problem:
If two self-heals are triggered on same inode in
parallel then one inode will be linked and the other
inode will not be linked as an inode with that gfid
is already linked in inode table. Calling inode-forget
on that inode leads to assert failure.
Fix:
Always use linked inode for performing self-heal.
Added inode-forgets in other places as well even though
its not really a memory leak.
Change-Id: Ib84bf080c8cb6a4243f66541ece587db28f9a052
BUG: 1091597
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/7567
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
Found the bug while browsing code. Very difficult to hit this though.
Change-Id: I84bd5d23d19b6aa16866fc0f1b56aa22cb8a5f1f
BUG: 1079215
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/7310
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
Gluster does not display performance.nfs.* options in help.
In Gluster NFS, write-behind is the only performance xlator
which gets loaded. Gluster volume set help should display all
the options provided by write-behind xlator.
Change-Id: I4a41151a6c15eeed8e8d123a6044c6f0c42b56b0
BUG: 1090826
Signed-off-by: Santosh Kumar Pradhan <spradhan@redhat.com>
Reviewed-on: http://review.gluster.org/7546
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
previously it was failed without showing any information why it is failed.
Now with this fix, when "." or any non alphanumeric character
present in volume name, it will give error messages
Change-Id: I17e8e69c08345c4d760f3ba333fe841e754bc9c8
BUG: 921215
Signed-off-by: ggarg <ggarg@redhat.com>
Reviewed-on: http://review.gluster.org/7364
Reviewed-by: Humble Devassy Chirammal <humble.devassy@gmail.com>
Reviewed-by: Atin Mukherjee <amukherj@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
1. Create directories in following fashion for history API's
usage when consumer is registered with libgfchangelog
shared library through gf_changelog_register.
scratch_dir/.history
scratch_dir/.history/.current
scratch_dir/.history/.processed
scratch_dir/.history/.processing
2. Added new file 'gf-history-changelog.c' and following APIs
are provided for consumers to process history changelogs.
1. gf_history_changelog_scan:
Move processed history changelog file from
.processing to .processed
2. gf_history_changelog_next_change:
Return the next history changelog file entry.
Zero means all history chanelogs are consumed.
3. gf_history_changelog_done:
Scan .processing directory and generate a list of
change entries.
4. gf_history_changelog_start_fresh:
For a set of changelogs, start from the begining.
NOTE: Though this patch provides above funcationalities.
It is considered functionally full with the
patch (http://review.gluster.org/#/c/6930/).
Change-Id: I200780c7278e0a6c008910d93faad5858a4b3e76
Original-author: Kotresh H R <khiremat@redhat.com>
Signed-off-by: Kotresh H R <khiremat@redhat.com>
Signed-off-by: Ajeet Jha <ajha@redhat.com>
Reviewed-on: http://review.gluster.org/6998
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Tested-by: Venky Shankar <vshankar@redhat.com>
During a peer-handshake, after the volumes have synced, and the list of
missed snapshots have synced, the node will perform the pending deletes
and restores on this list. At this point, the current snapshot list in
the node will be updated, and hence in case of conflicts arising during
snapshot handshake, the peer hosting the bricks will be given precedence
Likewise, if there will be a conflict, and both peers will be in the same
state, i.e either both would be hosting bricks or both would not be hosting
bricks, then a decision can't be taken and a peer-reject will happen.
glusterd_compare_and_update_snap() implements the following algorithm to
perform the above task:
Step 1: Start.
Step 2: Check if the peer is missing a delete on the said snap.
If yes, goto step 6.
Step 3: Check if there is a conflict between the peer's data and the
local snap. If no, goto step 5.
Step 4: As there is a conflict, check if both the peer and the local nodes
are hosting bricks. Based on the results perform the following:
Peer Hosts Bricks Local Node Hosts Bricks Action
Yes Yes Goto Step 7
No No Goto Step 7
Yes No Goto Step 8
No Yes Goto Step 6
Step 5: Check if the local node is missing the peer's data.
If yes, goto step 9.
Step 6: It's a no-op. Goto step 10
Step 7: Peer Reject. Goto step 10
Step 8: Delete local node's data.
Step 9: Accept Peer Data.
Step 10: Stop
Change-Id: I79be0f0f5f2a4f5c72277a4e77c2be732af432e1
BUG: 1061685
Signed-off-by: Avra Sengupta <asengupt@redhat.com>
Reviewed-on: http://review.gluster.org/7525
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com>
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
Reviewed-by: Atin Mukherjee <amukherj@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
During a glusterd handshake, a dictionary is passed among
the peers which contains, info of volumes, global opts,
and now also info of snaps and list of missed snaps
As it now contains more than just volume specific data,
renaming the dict in the code-base from "vols" to "peer_data"
Change-Id: Ib457172789ddd0d8978b08bceab0988c48e9eea7
BUG: 1061685
Signed-off-by: Avra Sengupta <asengupt@redhat.com>
Reviewed-on: http://review.gluster.org/7524
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
Reviewed-by: Atin Mukherjee <amukherj@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
The lvm snapshots of the bricks are mounted at /var/run/gluster/snaps/ or
/run/gluster/snaps. These paths being on a tempfs, on reboot are removed.
So when glusterd starts, we need to recreate these paths, activate the
respective logical volumes (lvm snapshots of the bricks), and mount
these logical volumes at their respective paths.
Change-Id: Ic5ef61e79a25d9830df717c592391965fe09db62
BUG: 1061685
Signed-off-by: Avra Sengupta <asengupt@redhat.com>
Reviewed-on: http://review.gluster.org/7452
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
After the drc cache get full, message "DRC failed to detect duplicates" keep
printed in the log.
The root cause is drc_compare_reqs use the wrong compare type. This function
should use type drc_cache_op_t as its input type. Since all rbtree related code
(except function rpcsvc_drc_lookup) in drc cache pass drc_cache_op_t as compare
type. Only rpcsvc_drc_lookup use type rpcsvc_request_t. It has been modified
too.
Change-Id: I925c097debe6b82f267986961fd4e7755f3de9af
BUG: 1089676
Signed-off-by: Yuan Ding <beback198611@gmail.com>
Reviewed-on: http://review.gluster.org/7519
Reviewed-by: Niels de Vos <ndevos@redhat.com>
Reviewed-by: Santosh Pradhan <spradhan@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
Persisting missing snapshot info on disk as well as in memory in
the following format:
-------------NODE-UUID--------------:--------------SNAP-UUID-------------=---------SNAP-VOL-ID------------:BRICKNUM:-------BRICKPATH--------:OPERATION:STATUS
927cb5fe-63da-48f5-82f6-e6a09ddc81c4:8258b18f-d408-483d-8239-204039dc6397=a17b4fe42c5a45f7a916438643edaa13: 3 :/brick/brick-dirs/brick3: 1 : 1
927cb5fe-63da-48f5-82f6-e6a09ddc81c4:8258b18f-d408-483d-8239-204039dc6397=a17b4fe42c5a45f7a916438643edaa13: 3 :/brick/brick-dirs/brick3: 3 : 1
927cb5fe-63da-48f5-82f6-e6a09ddc81c4:8258b18f-d408-483d-8239-204039dc6397=83a3cc05453b46b2a7eda4c9a9208638: 3 :/brick/brick-dirs/brick3: 1 : 1
This data will be stored on disk at /var/lib/glusterd/snaps/missed_snaps_list
In memory we maintain the data as a list of glusterd_missed_snap_info
in conf, the key for this list are the first two fields,
i.e NODE-UUID:SNAP-UUID.
For every NODE-UUID:SNAP-UUID, there can be multiple operations missed
on multiple bricks. So we maintain a list of glusterd_snap_op_t
for every node of glusterd_missed_snap_info
This list is maintained or updated during snapshot create, delete, and restore
operations which are the only operations that if missed, are recorded in this
list.
During snapshot create, if a node is down, or a brick is down, we don't
receive their mount point infos. snap_status of such bricks is marked as
-1, and their brick details are added to this list.
During snapshot delete, we check from originator node, if any other
nodes, holding bricks of the said snap are down. Those are also added to the list.
Also if the node is up, but the snapshot was pending for a snap
brick, and its snap_status is -1, we add that to the list too.
When a subsequent delete entry is processed for an already existing
create entry, we just mark the create entries status as done (2), and don't
add the delete entry to the list.
During snapshot restore, we check from originator node, if any other
nodes, holding bricks of the said snap are down. Those are also added to the list.
Also if the node is up, but the snapshot was pending for a snap
brick, and its snap_status is -1, we add that to the list too.
Like delete when a subsequent restore entry is processed for an already existing
create entry, we just mark the create entries status as done (2), and don't
add the restore entry to the list.
Change-Id: I54f63e28d3c40555d0f84528f38227103171f594
BUG: 1061685
Signed-off-by: Avra Sengupta <asengupt@redhat.com>
Reviewed-on: http://review.gluster.org/7454
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
gluster-devel@nongnu.org has moved to gluster-devel@gluster.org. All
occurrences in the current (non legacy) documentation and code have been
adjusted.
Change-Id: I053162e633f7ea14fd3eed239ded017df165147c
BUG: 1091705
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Reviewed-on: http://review.gluster.org/7573
Reviewed-by: Justin Clift <justin@gluster.org>
Reviewed-by: Anand Avati <avati@redhat.com>
Tested-by: Anand Avati <avati@redhat.com>
Instead of using 'int' for the sizes, use a 'size_t' as it is more
correct. Save the size of a fuse_dirent in a temporary variable so that
strlen() on the filename is called fewer times.
Also correcting some typos in comments.
Change-Id: Ic62d9d729a86a1a6a53ed1354fce153bac01d860
BUG: 1074023
Reported-by: Kaleb S. KEITHLEY <kkeithle@redhat.com>
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Reviewed-on: http://review.gluster.org/7547
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
Parallel builds failed due to make file would overrun
xdrgen (internally xdrgen uses tempfiles to add License
header).
Seperate out header and source generation and add explicit
dependency to fix it.
Change-Id: Id20f548493540b0f17a2300f0775646f9f20789c
BUG: 1090807
Signed-off-by: Harshavardhana <harsha@harshavardhana.net>
Reviewed-on: http://review.gluster.org/7572
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
In NLM callback path, if pmap_getport() FAILs, it just log the error
message saying "Is firewall running on the client". It may happen
that RPC services are not running e.g. "rpcbind" is not running or
nlockmgr (NLM) is not registered with portmapper which all can be
checked using "rpcinfo -p" command.
FIX:
Modify the log message to include the later case mentioned
above.
Change-Id: If422275b2ab59d1e974a6caa37132f31e9a34329
BUG: 1090782
Signed-off-by: Santosh Kumar Pradhan <spradhan@redhat.com>
Reviewed-on: http://review.gluster.org/7544
Reviewed-by: Niels de Vos <ndevos@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
* In the event of a DISCONNECT from a client, as part of cleanup,
entrylk objects are not removed from the blocked_locks list before
being unref'd and freed, causing the brick process to crash at
some point when the (now) stale object is accessed again in the list.
* Also during cleanup, it is pointless to try and grant lock to a
previously blocked entrylk (say L1) as part of releasing another
conflicting lock (L2), (which is a side-effect of L1 not being
deleted from blocked_locks list before grant_blocked_entry_locks()
in cleanup) if L1 is also associated with the DISCONNECTing client.
This patch fixes the problem.
Change-Id: I3d684c6bafc7e6db89ba68f0a2ed1dcb333791c6
BUG: 1089470
Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com>
Reviewed-on: http://review.gluster.org/7560
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
In a handshake, create a union of the missed_snap_lists of the two peers.
If an entry is present, its no op.
If an entry is pendng, and the peer entry is done, mark own entry as done.
If an entry is done, and the peer ertry is pending, its a no-op.
If its a new entry, add it.
Change-Id: Idbfa49cc34871631ba8c7c56d915666311024887
BUG: 1061685
Signed-off-by: Avra Sengupta <asengupt@redhat.com>
Reviewed-on: http://review.gluster.org/7453
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Rajesh Joseph <rjoseph@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
"9819fcedf10f1430d4969c86e6df4dfe975b7dcf" After the commit i
observed that we never used "hyper" as defined in .x previously
in the .[c,h] files.
Change-Id: I26152141bca6e789c4a3b3158fffe0a65e4b0878
BUG: 1090807
Signed-off-by: Harshavardhana <harsha@harshavardhana.net>
Reviewed-on: http://review.gluster.org/7566
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
Glusterfs build on RHEL 7 fails with "Installed (but unpackaged)
file(s) found". This patch fixes it.
BUG: 1058188
Change-Id: I8f37935b75ba02d085e75a66cdbc4017d82faa2e
Signed-off-by: Bala.FA <barumuga@redhat.com>
Reviewed-on: http://review.gluster.org/7557
Reviewed-by: Niels de Vos <ndevos@redhat.com>
Tested-by: Niels de Vos <ndevos@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Avoid modifying autogenerated files and keeping them in
repository - autogenerate them on demand from ".x" files
Change-Id: I2cdb1fe9b99768ceb80a8cb100fa00bd1d8fe2c6
BUG: 1090807
Signed-off-by: Harshavardhana <harsha@harshavardhana.net>
Reviewed-on: http://review.gluster.org/7526
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Niels de Vos <ndevos@redhat.com>
Reviewed-by: Anand Avati <avati@redhat.com>
This will change the following default configurables.
1. rollover-time: from 60 to 15.
2. fsync-interval: from 0 to 5.
Change-Id: I9c8db01376967c4f19547ec87f54833f8b139d29
Signed-off-by: Kotresh H R <khiremat@redhat.com>
Reviewed-on: http://review.gluster.org/7545
Reviewed-by: Venky Shankar <vshankar@redhat.com>
Tested-by: Venky Shankar <vshankar@redhat.com>
* In the event of a DISCONNECT from a client, as part of cleanup,
inodelk objects are not removed from the blocked_locks list before
being unref'd and freed, causing the brick process to crash at
some point when the (now) stale object is accessed again in the list.
* Also during cleanup, it is pointless to try and grant lock to a
previously blocked inodelk (say L1) as part of releasing another
conflicting lock (L2), (which is a side-effect of L1 not being
deleted from blocked_locks list before grant_blocked_inode_locks()
in cleanup) if L1 is also associated with the DISCONNECTing client.
This patch fixes the problem.
* Also, the codepath in cleanup of entrylks seems to be granting
blocked inodelks, when it should be attempting to grant blocked
entrylks, which is fixed in this patch.
Change-Id: I8493365c33020333b3f61aa15f505e4e7e6a9891
BUG: 1089470
Signed-off-by: Krutika Dhananjay <kdhananj@redhat.com>
Reviewed-on: http://review.gluster.org/7512
Reviewed-by: Raghavendra G <rgowdapp@redhat.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com>
Reviewed-by: Anand Avati <avati@redhat.com>
Problem:
When iozone is in progress, number of blocking inodelks sometimes becomes greater
than the threshold number of rpc requests allowed for that client
(RPCSVC_DEFAULT_OUTSTANDING_RPC_LIMIT). Subsequent requests from that client
will not be read until all the outstanding requests are processed and replied
to. But because no more requests are read from that client, unlocks on the
already granted locks will never come thus the number of outstanding requests
would never come down. This leads to a ping-timeout on the client.
Fix:
Do not account INODELK/ENTRYLK/LK for throttling
Change-Id: I59c6b54e7ec24ed7375ff977e817a9cb73469806
BUG: 1089470
Signed-off-by: Pranith Kumar K <pkarampu@redhat.com>
Reviewed-on: http://review.gluster.org/7531
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Krutika Dhananjay <kdhananj@redhat.com>
Reviewed-by: Anand Avati <avati@redhat.com>
It's a bit frustrating to run a test normally (e.g. during development),
see a failure, and then have to re-run it with DEBUG=1 to see what
actually failed. This tiny patch prints out the command line of the
test that just failed, even if DEBUG=0 (in fact only if DEBUG=0 since
otherwise it would be redundant).
Change-Id: Icffa096d0bcc6f35176f0e47d9d1bc538698ca2c
Signed-off-by: Jeff Darcy <jdarcy@redhat.com>
Reviewed-on: http://review.gluster.org/7537
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Anand Avati <avati@redhat.com>
RFE: Support wildcard in "nfs.rpc-auth-allow" and
"nfs.rpc-auth-reject". e.g.
*.redhat.com
192.168.1[1-5].*
192.168.1[1-5].*, *.redhat.com, 192.168.21.9
Along with wildcard, support for subnetwork or IP range e.g.
192.168.10.23/24
The option will be validated for following categories:
1) Anonymous i.e. "*"
2) Wildcard pattern i.e. string containing any ('*', '?', '[')
3) IPv4 address
4) IPv6 address
5) FQDN
6) subnetwork or IPv4 range
Currently this does not support IPv6 subnetwork.
Change-Id: Iac8caf5e490c8174d61111dad47fd547d4f67bf4
BUG: 1086097
Signed-off-by: Santosh Kumar Pradhan <spradhan@redhat.com>
Reviewed-on: http://review.gluster.org/7485
Reviewed-by: Poornima G <pgurusid@redhat.com>
Reviewed-by: Harshavardhana <harsha@harshavardhana.net>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
Problem : Reconfiguration of barrier timeout through gluster volume set shows a
success but it never changes the default timeout value which is 120 seconds.
After digging into the code deeper, it was found that timeout is never modified
in reconfigure() as the first check i.e. whether barrier is already enabled or
disabled always fails since barrier option is not modified in this request.
Fix : Introduced notify() in barrier translator which will take care of the rpc
request to enable/disable barrier. reconfigure() will simply set barrier
enable/disable and timeout options blindly without any validation.
Please note this patch only contains the changes in barrier translator however
from complete code flow perspective the caller in the glusterfsd mgmt should
call notify instead of reconfigure to fix this problem.
Change-Id: I1371b294935f6054da7c1dc6a9a19f1d861e60fb
BUG: 1085671
Signed-off-by: Atin Mukherjee <amukherj@redhat.com>
Reviewed-on: http://review.gluster.org/7428
Reviewed-by: Varun Shastry <vshastry@redhat.com>
Reviewed-by: Krishnan Parthasarathi <kparthas@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
brickinfo->brick_id was introduced to establish persistence of client xlator
names and AFR chanelog attributes (http://review.gluster.org/7155). The
snapshot volumes must also use the same IDs during snapshot create and
restore to maintain persistence.
Change-Id: I13d66d19b63520061ba9ec5f0ce661cf3b9eeafe
BUG: 1066778
Signed-off-by: Ravishankar N <ravishankar@redhat.com>
Reviewed-on: http://review.gluster.org/7477
Reviewed-by: Vijaikumar Mallikarjuna <vmallika@redhat.com>
Reviewed-by: Avra Sengupta <asengupt@redhat.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
In an environment with mixed architectures (32-bit servers, 64-bit
client), it is possible that the on-wire Reply on a READDIR(P) procedure
contains more direntries than the client can fit in the maximum size
that the fuse-request indicated.
A direntry is a dynamically sized structure, because the structure
contains the name of the entry. The client sends a maximum size in the
READDIR(P) Call to the server, and the server uses this size to limit
the number of direntries to return. In case the server can pack more
direntries in the requested maximum size (due to alignment differences
between the architectures), it can happen that the client unpacks the
list of direntries into a buffer that exceeds the maximum size that was
given in the initial fuse-request.
This change introduces a check for the maximum requested size of the
fuse-response in fuse_readdir_cbk() and fuse_readdirp_cbk(). When the
conversion from gluster-direntries to the fuse-direntry format takes
place, the maximum size is checked, and the 'extra' direntries are
discarded. The next readdir()/getdents() that is done, will fetch the
just discarded direntries again.
In addition to this bugfix, some extra logging in send_fuse_iov() and
send_fuse_data() has been added to help diagnosing similar issues.
Change-Id: If2eecfcdf9c248f3820035601446d2c89ff9d1a1
BUG: 1074023
Signed-off-by: Niels de Vos <ndevos@redhat.com>
Reviewed-on: http://review.gluster.org/7278
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Xavier Hernandez <xhernandez@datalab.es>
Reviewed-by: Anand Avati <avati@redhat.com>
A file descriptor like interface, backed by a string, on which
fprintf() like IO can be performed. Internally the backing string
is grown on demand.
Useful in generating virtual file content on the fly (used in meta)
Change-Id: I60d8751c4c750f3f06aa454a4ccd9909b3ac8ac7
BUG: 1089216
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/7508
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Introduce a new time format without "/" and spaces, in order to
be used in (virtual) filenames.
Change-Id: I468be54f1ec7f45265add4c458e19d95567439f7
BUG: 1089216
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/7507
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
Translators like meta, create virtual files with dynamic content
generated only at the time of open(). Therefore the file size
returned in lookup or stat is 0 (just like files in /proc). However
the VFS does not read beyond the size, and if the size is 0, no
READ ever reaches gluster for that file -- unless direct-io-mode
is enabled.
This patch allows translators to return "direct-io-mode" flag for
such 0-byte virtual files in xdata of open_cbk/create_cbk.
Change-Id: I3fe3312cd96baa4eecfe1247ab7255b4f455f049
BUG: 1089216
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/7506
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
the inode to be linked may have the last ref. loc_wipe() can
destroy it before inode_link() gets to ref it.
Change-Id: Ic2d44084e6e9c8289f35cae82c8e4575af105398
BUG: 1089216
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/7505
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
with members filled with pointers to appropriate default methods
Change-Id: I6cdc43e4f6776e2ad45cd5cbca5642e0c639ffde
BUG: 1089216
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/7504
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Pranith Kumar Karampuri <pkarampu@redhat.com>
In setattr, the inode times may have been explicitly set "back
in time". In such cases, if the inode ctx times are not force
set, then they continue to be higher and continue serving the
higher/older value in future calls to dht_inode_ctx_time_update()
Change-Id: I9cbfa7cf7c4069b0106d1f462de08c5d59bc91b5
BUG: 1083324
Signed-off-by: Anand Avati <avati@redhat.com>
Reviewed-on: http://review.gluster.org/7378
Reviewed-by: Harshavardhana <harsha@harshavardhana.net>
Tested-by: Harshavardhana <harsha@harshavardhana.net>
Tested-by: Gluster Build System <jenkins@build.gluster.com>
Reviewed-by: Raghavendra G <rgowdapp@redhat.com>
Reviewed-by: Vijay Bellur <vbellur@redhat.com>
